Construction and Building Materials 417 (2024) 135291

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

A study on the tensile fracture behavior of polypropylene fiber reinforced

concrete based on a microscale model
Huayi Wang a, 1, Xiongjun He a, Ming Zhou a, Bingyan Wei a, Weiwei Wu b, c, Guantao Zhou d,
Jia He e, *, 2
a
School of Transportation and Logistics Engineering, Wuhan University of Technology, Wuhan 430063, China
b
Department of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Kowloon 100872, Hong Kong
c
School of Engineering, The University of British Columbia, BC V1V 1V7, Canada
d
School of Science, Nanjing University of Science and Technology, Nanjing 210094, China
e
Department of the Built Environment, Eindhoven University of Technology, P.O. Box 513, 5600 MB Eindhoven, the Netherlands

A R T I C L E I N F O A B S T R A C T

Keywords: Polypropylene fibers are distributed in concrete with varying shapes and quantities, which is not conducive to
Mesostructure； Polypropylene fiber- microscopic modeling and analysis. This study aims to establish an effective and efficient crack calculation model
reinforced concrete； Interfacial Transition suitable for polypropylene fiber-reinforced concrete (PFRC) to reveal the reinforcement mechanism of poly­
Zone； Crack propagation； Finite element
propylene fiber-reinforced concrete. Firstly, from a microscale perspective, PFRC was regarded as a three-phase
analysis
heterogeneous material comprising aggregate, mortar (with embedded polypropylene fibers), and interfacial
transition zone (ITZ). Three-point bending and splitting tensile tests were conducted on mortar-aggregate
composite and mortar specimens with the fiber volume fraction as a variable. Furthermore, Peak stress ratio
and fracture energy ratio were introduced to characterize the modification effect of polypropylene fibers. The
influence of polypropylene fibers on the mechanical properties of mortar and ITZ was investigated, and the
modification mechanism of polypropylene fibers on concrete was analyzed from a microscale perspective.
Finally, a comprehensive microscale fiber-reinforced concrete cracking finite element model was established
based on the results of microscale experiments using two-dimensional image recognition and three-dimensional
numerical simulation methods. The research results show that PP fibers can significantly enhance the mechanical
properties of mortar, with flexural and tensile strength increased by 9.6–15.0% and 9.5–21.9%, and the
maximum fracture energy increased by 88.7%. The hydrophobicity of PP fibers limits the elimination of bubbles,
increases the porosity of ITZ, reduces the bonding strength of ITZ, and after maximum weakening, the flexural
strength, tensile strength, and fracture energy of ITZ are only 74.3%, 79.1%, and 37.6% of those without fiber
doping. The microanalysis model established based on this can effectively describe the relevant concrete tensile
fracture process indicators. The ratio of the stress peak calculated by the 3D model to the test is 0.997, while the
ratio of the stress peak calculated by the 2D model to the test is only 0.706. The simulation effect of the 3D model
is closer to reality. The influence of ITZ on the peak stress of concrete is more significant than that of mortar, and
improving the weakening effect of PP fibers on the bonding strength of ITZ is the key to improving the
macroscopic mechanical properties of PFRC.

1. Introduction alternative to conventional concrete, fiber-reinforced concrete (FRC) is a

multiphase heterogeneous cementitious material that utilizes fiber ma­
Concrete is the most widely used construction material, but ordinary terials as reinforcements. Compared to regular concrete (RC), FRC offers
concrete exhibits brittleness and low tensile strength. Under tensile higher ductility, strength, and durability [1]. Scholars have attempted to
stress, concrete is prone to cracking and difficult to repair. As an enhance the mechanical properties of concrete by using various fibers

* Corresponding author.
E-mail address: [email protected] (J. He).
1
https://orcid.org/0009-0006-9328-8235
2
https://orcid.org/0000-0002-2540-7873

https://doi.org/10.1016/j.conbuildmat.2024.135291
Received 3 October 2023; Received in revised form 2 December 2023; Accepted 30 January 2024
Available online 7 February 2024
0950-0618/© 2024 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
H. Wang et al. Construction and Building Materials 417 (2024) 135291

such as steel, polyvinyl alcohol (PVA), polyethylene (PE), and poly­ micron-sized fibers poses high computational demands on the computer
propylene (PP) [2–6]. Polypropylene fibers are considered a secondary regarding model generation, meshing, and calculations. Furthermore,
concrete reinforcement material. Due to the advantages of low cost, during the mixing, pouring, and compaction processes of concrete with
lightweight, easy blending, and significant improvement in concrete’s polypropylene fibers, the fibers are not uniformly distributed but rather
mechanical properties, polypropylene fiber reinforced concrete (PFRC) subjected to the compression of aggregates, rendering the method above
has been widely applied in key engineering projects such as buildings, unsuitable for simulating Polyolefin Fiber-Reinforced Concrete (PFRC)
bridges, and water conservancy [7]. from both a computational cost perspective and the irregular geometric
Existing research has demonstrated the enhancing effect of poly­ distribution of fibers. (2) The fiber and initial defects in the mortar are
propylene fibers on concrete performance. Microscopic analysis [8–10] considered part of the mortar, preparing a fiber-modified mortar matrix.
reveals that an appropriate dosage of polypropylene fibers can fill the The material parameters of the matrix and the interfacial transition zone
internal voids of concrete. The randomly distributed polypropylene fi­ (ITZ) are quantitatively measured through physical experiments. A
bers exert a restraining effect similar to a mesh-like structure on the three-phase microscopic concrete model is established: the mortar
concrete matrix, improving its crack resistance. On the other hand, both (fiber-modified), aggregates, and the interface transition zone (fiber-­
static and dynamic experiments on PFRC indicate that including an modified ITZ) between the fiber-modified mortar and aggregates.
appropriate volume fraction of polypropylene fibers can effectively Research has shown that [28] ITZ directly affects the macroscopic
enhance concrete’s tensile, flexural, and impact resistance properties mechanical properties of concrete. ITZ is a bonding material between
[11–13]. However, current experimental research on polypropylene aggregates and mortar composed of voids and cement, which is located
fiber-reinforced concrete mainly reflects the macroscopic behavior of inside the structure and has an irregular shape. Its thickness is usually
the material. Many scholars have explored the relationship between the between 0.01 to 0.05 mm [29]. It is not easy to study the mechanical
volume fraction of polypropylene fibers and flexural strength and tensile properties of ITZ directly. The usual practice is to define the strength of
strength, resulting in different findings and conclusions. The experi­ ITZ as 50% − 90% of mortar for trial calculation and compare macro­
mental results [10,14–21] show that when the fiber volume fraction is scopic results to determine specific values [30,31]. Although this
appropriate (about 0.2%− 0.4%), polypropylene fibers significantly method can achieve ideal simulation results, it cannot qualitatively
enhance the crack resistance of concrete. However, in most cases, when study ITZ. Hybrid specimens, as alternatives [32,33], are commonly
the volume fraction of polypropylene fibers exceeds the threshold 0.4%, used to determine the microstructure and bond strength of the ITZ [34].
the fibers have a negative effect on the overall performance of concrete. They have been employed to investigate the effects of aggregate
Scholars attribute the strengthening phenomenon to the stretching, roughness[35], mineral admixtures [36], and other factors on the ITZ.
sliding, and fracture of fiber fibers that enhance the performance of However, the composition of PFRC as a multiphase material is complex,
concrete. Scholars attribute the weakening phenomenon to the and different fiber volume ratios correspond to different interfacial in­
agglomeration phenomenon of high-content fibers without quantifying teractions between the aggregate and mortar in PFRC. The randomly
its inherent reasons. There is still insufficient research on the strength­ distributed micron-sized PP fibers in the ITZ necessitate a clear under­
ening mechanism of PFRC under low PP fiber content and the weakening standing of the influence of each phase material on the macroscopic
mechanism under high PP fiber content. Scholars attribute the mechanical properties of PFRC when studying the reinforcement
strengthening phenomenon to the stretching, sliding, and fracture of mechanism of polypropylene fibers. It is essential to investigate the
fiber fibers that enhance the performance of concrete. On the other fracture properties of the ITZ under the modification of PP fibers.
hand, scholars attribute the weakening phenomenon to the agglomera­ The finite element model links the microscopic analysis results of
tion phenomenon of high fiber content without quantifying its intrinsic concrete with the macroscopic mechanical properties. In the past few
reasons. There is still insufficient research on the strengthening mech­ decades, numerous mesoscale analysis models based on microstructure
anism of PFRC under low PP fiber content and the weakening mecha­ and local failure mechanisms have been proposed and effectively
nism under high PP fiber content. developed to analyze the influence of various component properties of
Currently, concrete research is mainly conducted from three per­ concrete on its macroscopic mechanical behavior. Schlangen et al.[37]
spectives: macro, micro, and micro [22–24]. Among them, the micro employ a discrete beam element-based equivalent frame structure to
model can simulate the mechanical properties of concrete and study its represent the continuous solid elements and utilize the mesh mapping
failure mechanism by separately simulating aggregates, mortar, and method to differentiate various concrete components. Removing the
interface transition zones (ITZ)[25]. On the microscopic level, failed beam elements simulates the initiation and propagation of con­
fiber-reinforced concrete is a multiphase composite material consisting crete cracks. Discrete element models [38] have also been utilized to
of aggregates, mortar, fibers, the interfacial transition zone (ITZ), and analyze mesoscale concrete models, but determining model parameters
initial defects (such as voids, microcracks, and air bubbles) [26]. is challenging. Based on the assumption of isotropic damage, the Con­
Therefore, in an ideal microscopic analysis model, all actual geometric crete Damage Model (CDP) [39,40] can consider stiffness degradation
characteristics of the microscopic structures (Including fibers with caused by plastic strain during tension and compression. It is commonly
disordered distribution and initial defects) should be accurately used to describe the mechanical response of concrete specimens under
considered and simulated. When considering discrete fibers separately, uniaxial tension and compression at the mesoscale level. However, it
the model consists of multiple random systems, including fibers, ag­ cannot accurately capture the crack propagation behavior. Furthermore,
gregates, and ITZ. The fibers should be randomly distributed among the the cohesive zone model (CZM) [34], based on the traction-separation
irregular aggregate particles and ensure they do not overlap with the law to represent the material’s fracture state, has been widely used for
aggregates while avoiding overlap between them. Covering all the simulating the initiation and propagation of cracks in concrete [41–46].
characteristics in a single random system would lead to modeling dif­ This model does not rely on high-precision experimental testing equip­
ficulties. Two main approaches exist for the microscopic simulation of ment and allows for quantitative analysis of the influence of various
existing fiber-reinforced concrete. (1) According to the method above mesoscale mechanical properties of concrete on its macroscopic
and the concrete mix design, only a few large-diameter fibers (primarily behavior. Therefore, this study uses the cohesive zone model (CZM) to
steel fibers) are incorporated [3,27]. This method is limited in terms of describe the crack patterns and macroscopic mechanical responses in
the types of fibers studied. For fibers with diameters in the micron range PFRC and verify the microscopic study’s accuracy.
(such as polypropylene fibers), due to their small diameter, the actual This paper aims to establish an effective and efficient cracking
situation often requires the inclusion of millions of fibers. In the study of calculation model applicable to polypropylene fiber reinforced concrete
microscopic models for fiber-reinforced concrete, the quantity of fibers (PFRC) and investigate the influence of material properties of various
directly determines the complexity of the model. Including many components of PFRC on its macroscopic mechanical performance. In the

2
H. Wang et al. Construction and Building Materials 417 (2024) 135291

Table 1
Sand performance index.
Material Apparent density Bulk density void ratio Broken value Mud content Needle-like granular content Stone powder content
(kg/m3 ) (kg/m3 ) (%) (%) (%) (%) (%)

Sand 2730 1698 38.0 Na 0 Na 5.4

Fig. 2. Appearance of polypropylene fibers.

Table 2
Mechanical properties of polypropylene fibers.
Fiber Length Diameter Tensile Density Elastic Elongation
(mm) (μm) strength (g/cm3 ) modulus (%)
Fig. 1. Fine aggregate size distribution curve.
(MPa) (MPa)

PP 12 32.7 469 0.91 4236 28.4

concrete tensile fracture calculation model, aggregates are usually
assumed to be non-rupturable [46], and the material parameters of
fiber-modified mortar can be tested using conventional methods. The characteristics of the sand as specified in the Technical Specifications for
effect of PP fiber on the Interfacial Transition Zone (ITZ) is studied using Highway Bridges and Culvert Construction (JTG/T3650–2020)[48] and
hybrid specimens. We selected PP fiber-modified mortar specimens and Fig. 1, the fineness modulus of the sand was 3.0, meeting the technical
PP fiber-modified mortar-aggregate hybrid specimens as the research requirements of Class II fine aggregate. The ART-JS polycarboxylic acid
objects. Firstly, three-point bending and splitting tensile tests were high-performance water reducer (Tianjin Jinxin Yuan) with a water
conducted on the mortar-aggregate hybrid and mortar specimens. Sec­ reduction rate of 27%, recommended for use by Concrete Admixtures
ondly, from the perspectives of load-deflection curves, interfacial (GB 8076–2008), was used. Tap water meeting the requirements of the
strength, and fracture energy, we quantitatively studied the fracture Standard for Mixing Water for Concrete (JGJ 63–2006) was used for
characteristics of the interface transition zone of the mortar and hybrid mixing. The appearance of polypropylene fibers is shown in Fig. 2.
specimens under PP fiber modification. Finally, based on the experi­ Determine the physical and mechanical properties of PP fibers according
mental results, a microscale finite element model for cracking to the Man-made fibre-Test method for tensile properties of staple fiber
fiber-reinforced concrete was established using two-dimensional image (GB/T 14337–2022)[49], and the results are shown in Table 2.
recognition and three-dimensional digital simulation methods. Na=Not applicated.

2. Materials and methods 2.2. 2.2 Specimen preparation

2.1. materials Tensile strength and Mode I fracture energy are critical parameters
for concrete fracture. This study conducted split tensile and three-point
This study used P.O 42.5 ordinary Portland cement (Huaxin Cement bending tests on polypropylene fiber-modified mortar specimens and
Co., Ltd., Enshi). The initial setting time of the cement was 210 min, and limestone-polypropylene fiber-modified mortar composite specimens to
the final setting time was 330 min. The density was 3.06 g/cm3 , and the investigate the ITZ’s tensile and bending fracture characteristics. The
specific surface area was 333 m2 /kg. All other indicators met the re­ effect of polypropylene fiber content on the composite specimens’ ten­
quirements of the Technical Specifications for Highway Bridge and sile strength and Mode I fracture energy was studied. Additionally, to
Culvert Construction (JTG/T3650–2020). Grade I fly ash was selected verify the reliability of the numerical analysis in this paper, a batch of
with a fineness of 8.2%, density of 2.54 g/cm3 , water demand ratio of cubic specimens of polypropylene fiber-modified concrete was prepared
92%, moisture content of 0.3%, SO3 content of 1.6%, and specific sur­ for relevant basic mechanical experiments. The limestone-mortar com­
face area of 460 m2 /kg. The rest of the indicators meet the technical posite beam specimens were composed of half mortar and half stone,
requirements of Class F Class I fly ash in Fly Ash Used in Cement and with the stone provided by Shandong Dashixiong Geotechnical Tech­
Concrete (GB/T 1596–2017). Grade S95 slag powder (Yichang Tengge) nology Company.
was chosen, and its technical specifications met the requirements of the According to relevant research [50–53] it can be concluded that the
Technical Specifications for Highway Bridge and Culvert Construction basic mechanical properties of PFRC reach the optimal state when the
(JTG/T3650–2020). volume ratio of polypropylene fibers (PP) is controlled at 0.22%
The fundamental performance indicators of fine aggregate sand − 0.40%; in order to systematically study the effect of PP fiber volume
(0–4.75 mm) were tested according to the Test Methods of Aggregate for ratio on the fracture performance of interface transition zone (ITZ), this
Highway Engineering (JTG E42–2005)[47]， and the test results are study designed four different fiber volume ratios of polypropylene fiber
shown in Table 1. According to the particle size distribution reinforced concrete (PFRC) based on their research. The

3
H. Wang et al. Construction and Building Materials 417 (2024) 135291

Table 3 for standard concrete cubic specimens. Firstly, the molds were coated
Mixing ratios of PFRC (kg/m3). with machine oil. The aggregates were weighed using an electronic
Cement Fly Sands Gravel Water Water Mineral Fiber platform scale and then batched into the HJW-60 single horizontal axis
ash reducer powder content forced concrete mixer for mixing. After pouring the concrete into the
400 40 677 1104 157 5.61 70 0(0%)/1.86 molds, they were placed on a vibrator for compaction. The entire process
(0.2%)/ of specimen production and curing is illustrated in Fig. 4.
3.72
(0.4%)/
5.58(0.6%) 2.3. Test methods

This experiment used the INSTRON 5967 universal material testing

mortar/concrete mix is shown in Table 3. machine and the MTS 810 testing machine to measure the fracture en­
The dimensions of the hybrid specimens for fracture energy testing ergy and tensile strength. A monotonically increasing displacement load
were 160 mm × 40 mm × 40 mm, with the mortar measuring at a constant rate of 0.02 mm/min was applied until the specimen failed.
80 mm × 40 mm × 40 mm and the limestone measuring The schematic diagram of specimen loading is shown in Fig. 5(a) and
80 mm × 40 mm × 40 mm. The natural interface had a height of (b), and the materials and equipment for the bending test are shown in
40 mm, including a 30 mm bonding surface and a 10 mm prefabricated Fig. 5(c) and (d). Fig. 5(e) and (f) show the materials and equipment for
notch, as shown in Fig. 3(a). To ensure the uniformity of the pre­ the tensile test. The load-deflection record is a continuous curve of P − δ,
fabricated notch, we used steel plates with specific thicknesses to make and the fracture energy GF is defined as the energy consumed per unit
grooves, as shown in Fig. 3(b). The dimensions of the hybrid specimens crack, which can be obtained by calculating the area under the P − δ
for tensile strength testing were 100 mm × 100 mm × 100 mm, with curve[34]。
the mortar measuring 100 mm × 100 mm × 50 mm and the limestone
measuring 100 mm × 100 mm × 50 mm, as shown in Fig. 3(c). Addi­ W = W0 + W1 + W2 + W3 (1)
tionally, beam specimens of pure mortar with polypropylene fiber in­ In the formula, W0 is the energy generated by the load P(δ) ;W0 =
clusion were fabricated with dimensions of 160 mm × 40 mm × 40 mm ∫ δmax
P(δ)dδ;W1 + W2 represents the gravity work of the experimental
and 100 mm × 100 mm × 100 mm for comparison. 0

During the casting of the hybrid specimens, ensuring sufficient beam between two supports; W3 represents the work done by the gravity
moisture on the surface of the limestone was crucial. The limestone was of the loading head. In this study, the loading head and device are in­
wholly submerged in water for 24 h before being taken out and used tegrated so that this part can be ignored.
within half an hour for mortar pouring. The bonding surface was kept For mortar specimens, the fracture process of the experimental beam
moist throughout the pouring process but without visible water. The consists of two parts: W1 ≈ W2 = 12 mgδmax ; The fracture energy GFM of
purpose of the hybrid specimens was to investigate the influence of the mortar is:
polypropylene fiber parameters on the ITZ performance. The fabrication ∫ δmax
W P(δ)dδ + mgδmax
method involved placing the wet limestone on one side of the mold, GFM = = 0 (2)
A A
pouring mortar on the other, and inserting steel plates for the pre­
fabricated cracks. After compaction using a vibrating table, the speci­ In the formula, mg is the weight of the specimen between two sup­
mens were left indoors to cure for 24 h and then de-molded and cured ports; δmax is the maximum deflection at the mid-span when the spec­
for 28 days in a standard concrete curing chamber at a temperature of imen fails; A is the area of the beam fracture range, A = b(h − a).
20 ± 2 ◦ C and a humidity of 98%. For the hybrid specimen, the gravity work done by the mortar and
Moreover, for experimental validation, the specimens were prepared rock from the two supports of the fracture process experimental beam to
according to the requirements of GB/T 50081–2019 [54] and GB/T the prefabricated crack surface is not equal, W1 = 12m1 gδmax , W2 =
50082–2009 [55], with dimensions of 150 mm × 150 mm × 150 mm 1
2m2 gδmax , the ITZ fracture energy GFI is:

Fig. 3. Detailed dimensions of the specimen.

4
H. Wang et al. Construction and Building Materials 417 (2024) 135291

Fig. 4. The process of production and maintenance of specimens.

∫ δmax
W P(δ)dδ + 12m1 gδmax + 12m2 gδmax between the cement matrix and aggregate particles in mortar, thereby
GFI = = 0
(3) enhancing the displacement response of mortar specimens with an in­
A A
crease in the volume fraction of polypropylene fiber. However, for the
In the formula, m1 g represents the weight of mortar from one part of Hybrid specimens, including polypropylene fiber significantly weakens
the support to the precast crack surface, and m2 g represents the weight the displacement response and peak load of the Interface Transition
of rock from the other part of the support to the precast crack surface. Zone (ITZ) between Mortar and limestone. The weakening effect be­
The tensile strength is: comes more pronounced with a higher volume fraction of polypropylene
2P P fiber.
Tm = = 0.637 (4)
πA C2
2.4.2. Flexural and splitting tensile strengths
In the formula,Tm is the tensile strength of the specimen, MPa; P is
The bending strength and splitting tensile strength test results of
the maximum load, N;Cis the side length of the cube, mm. GB/T
mortar specimens and Hybrid specimens are shown in Fig. 7. The
50081–2019 [54] stipulates that the splitting tensile strength measured
parameter mean results are shown in Table 4. The experimental results
using non-standard specimens (100 mm × 100 mm × 100 mm) should
of polypropylene fiber modification indicate that polypropylene fibers
be multiplied by a size conversion factor of 0.85.
enhance mortar specimens’ splitting tensile strength and flexural
strength within a specific range. Compared to the unfired mortar, the
2.4. Test results flexural strength of mortar with 0.2%, 0.4%, and 0.6% volume fractions
of polypropylene fibers improves by 9.6%, 15.0%, and 11.1%, respec­
2.4.1. Load-deflection response tively, after 28 days. Furthermore, the splitting tensile strength increases
According to the aforementioned experimental method, the load- by 9.5%, 21.9%, and 10.7% for the corresponding volume fractions. The
displacement curves of Mortar and Hybrid specimens with four inclusion of polypropylene fibers has a significant effect on the flexural
different fiber volume fractions were plotted as shown in Fig. 6. The and splitting tensile strength of mortar. However, this enhancement
specimen labels start with the letter "P" to represent Polypropylene fiber, effect is not unlimited. The flexural and splitting tensile strengths of the
followed by a numerical value indicating the volume fraction of poly­ mortar with a 0.6% volume fraction of polypropylene fibers are slightly
propylene fiber. For example, P-0.2 signifies a volume fraction of 0.2% lower than those with a 0.4% volume fraction.
for polypropylene fiber. The results indicate that, for both the Mortar On the other hand, the test results for the flexural strength and
and Hybrid specimens, the descending segment of the load-displacement splitting tensile strength of Hybrid specimens show that an increase in
curve is vertical, highlighting the brittle failure characteristics of low- the volume fraction of polypropylene fibers leads to a significant
parameter polypropylene fiber-modified mortar and the Interface reduction in the mechanical performance of the ITZ. For instance,
Transition Zone (ITZ). Adding polypropylene fiber improves the bond

5
H. Wang et al. Construction and Building Materials 417 (2024) 135291

Fig. 5. Schematic diagram of specimen loading.

Fig. 6. Flexural load-deflection curve.

6
H. Wang et al. Construction and Building Materials 417 (2024) 135291

Fig. 7. Variation of breaking strength with fiber volume ratio.

0.4%, and 0.6% polypropylene fiber content increased by 18.1%, 66.0%,

Table 4 and 88.7% at 28 days, respectively. It can be seen that adding poly­
Specimen fracture parameters. propylene fiber improves the fracture energy of the mortar. However,
Specimen Flexural strength Splitting tensile strength Fracture energy the fracture energy results from the Hybrid specimens reveal that the
(MPa) (MPa) (N/mm) presence of polypropylene fibers severely reduces the fracture energy of
Mortar Hybrid Mortar Hybrid Mortar Hybrid the ITZ. For instance, compared to the unfired mortar, the fracture en­
ergy of Hybrid specimens with 0.2%, 0.4%, and 0.6% volume fractions
P-0 8.80 4.74 4.02 2.32 0.169 0.049
P-0.2 9.64 4.50 4.40 2.15 0.199 0.038
of polypropylene fibers is only 76.5%, 58.0%, and 37.6%, respectively,
P-0.4 10.12 4.00 4.90 2.02 0.280 0.029 of the unfired strength after 28 days.
P-0.6 9.78 3.52 4.45 1.78 0.319 0.019
2.4.4. Result analysis
In order to further analyze the reinforcing effect of PP fiber on
mortar, electron microscopy scanning of polypropylene fiber reinforced
mortar is shown in Fig. 9.
From Fig. 9. It can be seen that there is a disorderly distribution of
tiny polypropylene fibers in the polypropylene fiber-reinforced mortar.
During the cracking process of the matrix, the tensile strength of poly­
propylene fibers is much greater than that of mortar. The elongation,
slip, fracture, and detachment of disorderly distributed fibers under
tensile stress increase the energy required for matrix crack propagation,
thereby improving the crack resistance of mortar. The macroscopic
mechanical behavior manifests as the addition of PP fibers increasing
the mortar’s tensile strength and fracture energy.
Perform image processing on the location of mortar fracture in
Hybrid specimens. The void ratio of the ITZ is plotted in Fig. 10. It can be
observed that the inclusion of polypropylene fibers significantly in­
creases the void ratio of the ITZ. In concrete, due to the capillary action,
free water inside the mortar always flows toward the surface of aggre­
gates during the vibration process [35]. This capillary action leads to the
Fig. 8. Variation of fracture energy with fiber volume ratio. formation of numerous water pockets and voids on the aggregate sur­
face. Since polypropylene fibers are hydrophobic [56], they restrict
water ingress into the concrete and simultaneously attract more air
compared to the unfibered mortar, the flexural strength of Hybrid
bubbles to adhere to their surface, limiting the exclusion of bubbles. In
specimens with 0.2%, 0.4%, and 0.6% volume fractions of poly­
Fig. 10, it is evident from the measurement of the Hybrid specimen
propylene fibers are only 94.7%, 84.4%, and 74.3% of the unfibered
mortar that the size of the voids increases with an increase in the fiber
version after 28 days, respectively. Similarly, the splitting tensile
volume fraction. Adding PP fibers increases ITZ’s local porosity and
strength is only 95.6%, 89.8%, and 79.1% of the unfibered version for
reduces the bonding area between mortar and aggregate. The macro­
the corresponding volume fractions. These results indicate that adding
scopic mechanical properties show a decrease in the bonding strength
polypropylene fibers weakens the strength of the section transition zone.
and fracture energy of the interface between aggregate and mortar
matrix, which damages the mechanical properties of ITZ.
2.4.3. Fracture energy
Therefore, polypropylene fiber-modified concrete cannot achieve the
The fracture energy of the mortar and ITZ, calculated based on the
optimal modification effect by increasing the fiber volume fraction. It is
relevant formulas, is presented in Fig. 8. The average values of fracture
also necessary to pay attention to the influence of polypropylene fiber on
energy for the mortar and ITZ parameters are summarized in Table 4.
the strength of aggregates and ITZ. It is necessary to establish relevant
The experimental results demonstrate that compared with the mortar
calculation models to study the balanced relationship between the
without fiber, the fracture energy of the mortar mixture with 0.2%,
reinforcement effect of polypropylene fiber on mortar and the interface

7
H. Wang et al. Construction and Building Materials 417 (2024) 135291

Fig. 9. Scanning electron microscopy of polypropylene fiber reinforced mortar.

Fig. 11. Polyhedral aggregate interference judgment.

vibration processes due to the compression of aggregates, making it

more expensive to model them separately. Therefore, this study con­
siders polypropylene fiber as a part of mortar and establishes a three-
phase microscopic concrete model: mortar, aggregate, and the inter­
face transition zone (ITZ) between mortar and aggregate. Most methods
Fig. 10. Variation of porosity with fiber volume ratio.
for generating random polyhedral aggregate models are similar to [62,
63]： (1) generating polyhedral aggregates; (2) randomly generating
weakening effect. aggregate center point coordinates within the matrix range; (3) placing
aggregates according to the center point coordinates; (4) determining
3. 3D mesostructure model of PFRC whether there is interference between the later placed aggregates and
the previously placed aggregates; (5) repeating steps 2–4 until the vol­
3.1. Establishment method of PFRC microscopic modeling ume rate is qualified and terminating. The key to the above steps lies in
step (4), where the most common method requires that the distance
There are two standard methods for mesoscopic modeling of con­ from the generated aggregate center point coordinate to the previously
crete. The first method involves scanning actual concrete cubes using an generated aggregate center point coordinate is greater than the sum of
X-ray computed tomography (X-CT) scanner and generating a mesh the radii of the two circumscribed spheres. However, this method’s
model from the digital scans[45,57,58]. Li et al.[59] used a CT scanner volume fraction of the aggregate generated is relatively low. Pouria’s
to generate specimens and obtain the final microstructure. However, [64] study showed that when the volume fraction of aggregates reaches
this method cannot accurately identify and distinguish the boundaries of 35%, there will be intersections between aggregates. The above method
aggregates, and some densely packed regions result in overlapping ge­ is not suitable for generating aggregates with high-volume fractions.
ometries, requiring manual processing of the mesh model. Moreover, Zhou[62] proposed a new method for discriminating interference based
this method involves multiple steps, such as specimen casting, curing, on Naderi’s [27] research on fiber intersection. When generating a
X-CT scanning, and model reconstruction, which are costly and random polyhedron, vertex coordinate data is stored, and the minimum
time-consuming. Another approach was proposed by Xie et al.[60], who distance dmin between all vertices of two aggregates is calculated, such as
utilized laser scanning to capture the natural shape of aggregates and Eq.(5). The schematic diagram of interference judgment is shown in
establish a database. The point lattice method was then employed for Fig. 11[62].
placement. Although this method can reflect the actual conditions of ( ( ))
aggregates, it is difficult to control the irregularity of aggregates. The dmin = min dis AVCtp , AVCtq > l0
(5)
second method uses relevant algorithms to generate polyhedral aggre­ t ∈ [1, n], p, q ∈ [1, N]
gates and place them in the corresponding positions. The standard
method for generating polyhedral aggregates is randomly selecting a AVCtp is the vertex coordinate of the generated aggregate p, AVCtq is
certain number of vertices on the sphere and then connecting the plane the vertex coordinate of the following generated aggregate q, n is the
topology to form a polyhedron [61]. number of vertices of the polyhedral aggregate section, and N is the
Table 2 shows that the length of polypropylene fibers is 12 mm, and number of aggregates. When the number of aggregate vertices is large
the diameter is 32.7μm. Due to the low elastic modulus and soft texture enough and l0 = 0, all vertices do not come into contact, and no pene­
of polypropylene fibers, polypropylene fibers are not distributed in a tration of aggregate occurs. However, it is not certain that the number of
single shape inside the concrete during the mixing, pouring, and aggregate vertices is ideal when generating the aggregate. The l0 value

8
H. Wang et al. Construction and Building Materials 417 (2024) 135291

Table 5 inside the material. The cohesive zone model can simulate the entire
Particle size distribution of aggregates in concrete. crack initiation, evolution, and failure process for brittle materials such
Sieve size (mm) Total percentage passing (%) Total percentage retained (%) as concrete and quasi-brittle materials. This study’s nonlinear mechan­
ical response of mortar and the interface transition zone (ITZ) in the
12.7 0 100
9.5 21.93 78.07 micro-scale model are characterized using cohesive force models [43,45,
4.75 75.82 24.18 65].
2.36 97.79 2.21 Cohesive elements are thin elements embedded between adjacent
meshes, which can be used to simulate mortar cracking between mortar
elements or the fragmentation of aggregates between aggregate ele­
ments [44], as shown in Fig. 14(a). They can also be inserted at the
mortar-aggregate interface to simulate weak defect locations, such as
the Interface Transition Zone (ITZ) in concrete, as depicted in Fig. 14(b).
The specific implementation method in the commercial software ABA­
QUS involves exporting the inp file, reading and numbering the element
nodes, inserting cohesive elements between shared nodes, and then
renumbering all nodes before writing them back into the inp file to
establish a finite element model with cohesive interface elements.
Establish a finite element model of the 3D microscopic concrete
cohesive interface element (CIE), as shown in Fig. 15. Firstly, different
sets are defined for the micro-scale concrete model structure: aggregates
and mortar. Then, the mesh is generated as depicted in Fig. 15(a). Next,
three sets of cohesive elements, namely CZM_AGG (Aggregate),
CZM_MOR (Mortar), and CZM_ITZ (Interfacial Transition Zone), are
inserted into the initial mesh, as shown in Fig. 15(b), (c), and (d).
Finally, different traction-separation softening laws are assigned to
simulate concrete cracking. The same approach applies to the two-
dimensional micro-scale model, as shown in Fig. 16.

3.3. Cohesive constitutive and material properties

Fig. 12. Mesoscopic model of polyhedral aggregate with a volume fraction
of 30%. A bilinear traction-separation relationship applies to quasi-brittle
materials such as cement mortar and aggregates [66,67]. The constitu­
tive responses of the cohesive models under pure tensile, compressive,
can be appropriately increased to reduce the possibility of aggregate
and shear stress conditions are shown in Fig. 17. Fig. 17(a) represents
penetration.
the primary Type I bilinear cohesive model, which assumes that the
The material models were generated using the method proposed by
separation of the material interface is controlled by displacements
Zhou et al.[62] in this study. The particle size distribution of the ma­
perpendicular to the interface. Fig. 17(b) depicts the primary Type II
terials is presented in Table 5， and a micro-scale model was estab­
bilinear model, which assumes that the separation of the material
lished, as shown in Fig. 12. Meanwhile, the method proposed by Li et al.
interface is related to displacements tangential to the interface.
[34] was selected for two-dimensional micro-scale modeling to conduct
Before the stress exceeds the damage strength, the cohesive zone
comparative analyses. After cutting the concrete specimen, the cut
model exhibits linear elastic constitutive law:
surface was captured using a digital camera, as shown in Fig. 13(a).
Subsequently, image processing techniques were employed to determine tn = K 0n 〈δn 〉 (6)
the boundaries of the particles, as depicted in Fig. 13(b). After
enhancing the processed images, a finite element computational model ts = K 0s δs (7)
was obtained and imported into ABAQUS, as illustrated in Fig. 13(c).
In the equation, tn and ts represent the normal stress and shear stress
of the cohesive element, respectively;〈〉is the Macaulay bracket， which
3.2. Generation of cohesive zone model is 0 in compressed state; δn and δs represent forward opening and shear
slip displacement; K0n (K0s ) represents the penalty stiffness of the cohesive
A topological connection exists between cohesive and finite ele­ element.
ments, but their constitutive relationships are independent. According When the stress of the cohesive element meets the damage initiation
to the cohesive zone model theory, this model can effectively describe criterion and enters the damage evolution stage, this study adopts the
the failure and damage caused by crack initiation and propagation quadratic nominal stress criterion, which can be expressed as:

Fig. 13. Two-dimensional finite element modeling of concrete.

9
H. Wang et al. Construction and Building Materials 417 (2024) 135291

Fig. 14. Two types of cohesive elements in the concrete meso-cracking model.

Fig. 15. Three-dimensional concrete models with different cohesive interface elements.

Fig. 16. Two-dimensional concrete models with different cohesive interface elements.

{
〈tn 〉
}2 { }2 { }2
ts tt For bilinear models, D can be represented as[68]:
+ + =1 (8) ( )
tn0 ts0 tt0 δf δmax − δ0m
D = mmax (m f ) (10)
In the equation, tn , ts and tt represent the normal stress and two δm δm − δ0m
tangential stresses of the cohesive element, respectively; tn0 ,ts0 (tt0 )
In the equation, δmax m is the maximum effective relative displacement
represent the tensile strength and shear strength.
When the cohesive element is damaged, the stiffness can be obtained during the loading process, δ0m and δfm are the effective relative
expressed as[68]： displacements at the onset of damage and final failure, respectively. δ0m
( ) can describe the damage evolution of cohesive elements under the
K = 1 − D K0 (9) combined action of normal and tangential deformation, and its expres­
In the formula, K represents the damage stiffness; D represents the sion is [43]：
stiffness damage factor,D = 0represents no damage，D = 1represents √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
complete failure of the element. δ0m = 〈δn 〉2 + δs 2 + δt 2 (11)

10
H. Wang et al. Construction and Building Materials 417 (2024) 135291

Fig. 17. Constitutive response of cohesive elements.

4. Simulation results and discussion

Table 6
Material properties for aggregate, mortar, and ITZ.
Based on the cracking characteristics of concrete microstructure, the
Material parameters Bulk elements Cohesive elements failure mode of PFRC can be observed. By comparing the calculation
Aggregate Mortar CIE_AGG CIE_MOR CIE_ITZ results of the microstructure model with the experimental results of
Young’s modulus, E 72 26 - - - PFRC specimens, the rationality of the experimental results of fiber-
(GPa) modified mortar and interface can be verified. A cubic specimen of
Poisson’s ratio，ν 0.16 0.2 - - - three-dimensional concrete is constructed with dimensions of
Density, ρ(kg/m3 ) 2700 2200 2700 2200 2000 50 × 50 × 50 mm. The left surface of the model is fully constrained,
Normal elastic - - 106 106 106
while the right surface is subjected to a uniformly distributed displace­
stiffness, Kn
(GPa/m) ment. By connecting the right surface to a reference point, an axial
Tangential elastic - - 106 106 106 displacement is applied to the reference point. The final displacement is
stiffness, Ks set at 0.1 mm[74]. The reference point’s reaction force and displace­
(GPa/m) ment curves are recorded and combined. The equivalent stress result for
Material coefficient of - - 1.2 1.2 1.2
the BK criterion, η
the three-dimensional concrete model is determined by dividing the
node reaction force by the cross-sectional area. For the two-dimensional
concrete model, the equivalent stress result is obtained by dividing the
The determination of δfm is relatively complex, which reflects the node reaction force by the length of the coupling edge. Additionally, the
fracture energy release rate of the element under mixed stress state. B-K equivalent strain of the concrete is expressed as the ratio of the
（Benzeggagh-Kenane）[69] proposed a fracture criterion suitable for displacement of the reference point to the side length of the model in the
describing quasi-brittle materials from an energy perspective, and δfm can direction of elongation. Use the ABAQUS/Explicit solver for quasi-static
be expressed as: calculation analysis, and select "smooth analysis" for displacement
[ ( )( )η ] application. The ABAQUS/Explicit solver does not require iteration and
2
δfm = 0 0 GI + GII − GI
Gs
(12) can solve highly nonlinear equations effectively. It is worth noting that
K δm Gn + Gs selecting a quasi-static loading method requires considering the effects
of loading time, mesh density, and other factors on the accuracy of the
In the formula, GI and GII represent the tensile (Type I) and shear
calculation results.
(Type II) fracture energies, respectively;Gn and Gs represent the work
done by tensile and shear forces, respectively;ηis the BK criterion
correction parameter. 4.1. loading time
Due to the limitations of experimental conditions, we cannot deter­
mine all necessary material parameters. However, the inherent proper­ Existing research usually judges the rationality of a quasi-static state
ties of materials, such as elastic modulus, Poisson’s ratio, and density, based on energy. The quasi-static state requires that the total kinetic
can be found in references [46,70], and their variations are minor, so energy of the deformed material should not exceed 5% − 10% of the
they can be ignored. The tensile strength and Type I fracture energy are total internal energy [75]. The total kinetic energy of the model is
shown in Table 4. Research on concrete shear experiments [71–73] usually positively correlated with the displacement loading rate. In this
shows that the Type II fracture energy is approximately 20–25 times that study, the selected loading displacement is a fixed value of 0.1 mm. In
of Type I fracture energy. Therefore, this study uses this proportional order to obtain accurate calculation results, it is necessary to choose an
relationship to define the Type II fracture energy. Other relevant appropriate loading rate and set the corresponding load step time in the
calculation parameters are shown in Table 6。 model. In this study, load step times were set at 0.001 s, 0.0025 s,
0.005 s, 0.0075 s, and 0.01 s, respectively, corresponding to loading
rates of 100.00 mm/s, 40.00 mm/s, 20.00 mm/s, 13.33.00 mm/s, and
10.00 mm/s.

11
H. Wang et al. Construction and Building Materials 417 (2024) 135291

reflecting the crack propagation law. In the experiment, the minimum

coarse particle size of the sand was 2.36 mm, so the maximum grid size
selected for this study was 2 mm, which ensured that sand with a par­
ticle size greater than 2.36 mm could be reasonably divided into grids.
At the same time, grid sizes of 1.7 mm and 1.5 mm were set for com­
parison, as shown in Fig. 19. Mesh A has 257466 solid elements and
511182 cohesive elements, while Mesh B and C have 342260, 679474,
and 442008, 877482, respectively.
The impact of grid density on the calculation results is shown in
Fig. 20, where Fig. 20(a) shows the macroscopic stress-strain curve
under different densities, Fig. 20(b) is the energy dissipation strain
curve, which shows that the dependence of the stress-strain and energy
dissipation strain curves on the grid density under the selected element
size can be ignored. However, during the calculation, the growth rate
increased significantly. In the later calculation, a grid size of 2 mm
(Mesh A) was selected for all grids.
Fig. 18. Comparison of stress-displacement curves under different
loading rates.
4.3. Calculation result
As shown in Fig. 18, the peak stress tends to stabilize as the loading
rate decreases, and when the loading rate is less than 20 mm/s (corre­ The mechanical properties of polypropylene fiber-modified mortar
sponding to a loading step time of 0.005 s), the curve is unchanged, but and ITZ were measured in the second section of the experiment and
the computational time increases with the loading step time. Therefore, incorporated into the 3D (2D) model established above. This study as­
all model load step times were taken as 0.005 s, and a displacement of sumes that the damage index D ≥ 0.99 is a wholly damaged crack, and
0.1 mm was loaded. the crack paths of PFRC under different fiber volume fractions are
calculated as shown in Table 7-Table 8.
Although there are some differences between the 3D model and the
4.2. Mesh density 2D model, there are similarities in crack morphology and failure
morphology in the three states (At peak stress, Softening range, and
For quasi-static analysis, the influence of grid size on the accuracy of Failure):(a) In the early stage of loading, the specimen is in the elastic
the calculation results cannot be ignored. An overly detailed grid will stage, and no cracks are generated; (b) After continuous loading for a
increase unnecessary time costs, and a rough grid will lead to non- period of time, a small number of micro-cracks randomly appear in the
convergence of the calculation results and difficulty accurately interface transition zone (ITZ) between mortar and aggregate, and are

Fig. 19. Computational models for different mesh densities.

Fig. 20. Effect of mesh density.

12
H. Wang et al. Construction and Building Materials 417 (2024) 135291

Table 7
Fracture of two-dimensional specimens.

dispersed throughout the entire model. Due to the weaker mechanical We can conclude from the figure that, using the experimental data in
properties of ITZ compared to mortar and the uneven aggregate surface, Section 2.4, the 3D model calculation results are relatively close to the
stress concentration is prone to occur near the transition zone with observed values regarding stress-strain trends and load peaks. However,
mortar. Therefore, cracks first originate here. The larger the volume although the stress-strain curve trend of the 2D model calculation results
ratio of PP fibers, the weaker the ITZ performance and the more cracks is similar to the experimental values, the stress peaks differ significantly.
are generated; (c) As the tensile displacement further increases, the re­ The average ratio between the calculated results of the 3D model and the
action load reaches its peak, and many micro-cracks will be generated in experimental results is 0.997, while the average ratio between the
the interface transition zone (ITZ) between mortar and aggregate. calculated results of the 2D model and the experimental results is only
Cracks near large-volume aggregates will extend from the initiation 0.706, which is much smaller than the experimental test results, indi­
position of ITZ to mortar; (d) In the subsequent stage, the reaction force cating that the computational performance of a 3D computational model
on the specimen gradually decreases with the increase of tensile is more accurate than that of a 2D model.
displacement, and cracks near large-volume aggregates further develop The results in Section 2.4 indicate that adding polypropylene fibers
and avoid the aggregates. Some cracks merge into dominant cracks, has the best reinforcement effect on the tensile strength of mortar at a
while cracks near small-volume aggregates stop expanding due to stress fiber volume fraction of 0.4%, and the reinforcement effect is slightly
redistribution; (e) In the end, the dominant cracks near the large-volume lower than 0.4% at a fiber volume fraction of 0.6%. The effect of poly­
aggregate continued to expand, and multiple dominant cracks merged to propylene fibers on the fracture energy of mortar has been the rein­
form the main crack that penetrated the entire model, causing the forcement effect until the fiber volume ratio is 0.6%, and on the other
specimen to fail. hand, it weakens both the tensile strength and fracture energy of ITZ.
The stress-strain curves and experimental test results of PFRC with The stress peak calculated using the 3D model of this law is consis­
different fiber volume fractions are calculated using the 3D and 2D tent with the experimental results: the ultimate tensile strength of PFRC
models for further analysis. These results are shown in Fig. 21. The peak with a fiber volume ratio of 0.4% has the most significant strengthening
stress results are depicted in Fig. 22(a), while Fig. 22(b) compares the effect compared to RC. However, when the fiber volume ratio exceeds
calculations from the 3D and 2D models and the experimental results. 0.4%, it weakens the tensile strength of PFRC. The combined result of

13
H. Wang et al. Construction and Building Materials 417 (2024) 135291

Table 8
Fracture of three-dimensional specimens.

Fig. 21. Macroscopic stress-strain curves of concrete in tension.

the decrease in tensile strength of mortar and ITZ causes this weakening The calculation results of the 2D model show different patterns from
effect. The peak stress of concrete is more sensitive to the tensile the experiment: the ultimate tensile strength of PFRC with a fiber vol­
strength of mortar and ITZ than fracture energy. ume ratio of 0.2% is slightly enhanced compared to RC, and an increase

14
H. Wang et al. Construction and Building Materials 417 (2024) 135291

Fig. 22. Peak tensile stress analysis.

models, the sensitivity analysis of mortar and ITZ tensile strength in

different dimensions was conducted based on the mechanical properties
of fiber-free materials, as shown in Fig. 23.
By changing the tensile strength of mortar (3.5 ± 0.5 MPa) and ITZ
(2.0 ± 0.5 MPa), we can observe that whether it is a 3D model or a 2D
model, the ITZ strength has a significant impact on the peak variation of
concrete tensile strength. Therefore, when we focus on the improvement
effect of various materials on the tensile strength of concrete materials,
we should pay attention to the influence of fibers on the interface
transition zone. Increasing the strength of mortar in 3D models can
slightly improve the peak tensile strength of concrete, but enhancing the
strength of mortar in 2D models does not seem to have a significant
effect on the peak tensile strength of concrete. It can be explained that
the ratio of the cracked surface to volume in a three-dimensional model
is greater than the ratio of the cracked line segment to area in a two-
dimensional model. The three-dimensional model improves the
constraint in the thickness direction, and the simulation effect is closer
to the actual situation.
In order to analyze the fracture process of the 3D model in more
Fig. 23. Tensile strength sensitivity analysis. detail, the energy consumption and crack area as a function of strain are
plotted in Fig. 24.
in fiber volume ratio exceeding 0.2% does not increase in PFRC tensile As shown in Fig. 24 (a), The fracture of the specimen consumes the
strength. dissipated fracture energy. The dissipated fracture energy strain curve
To analyze the computational differences between 2D and 3D can be roughly divided into three stages: (a) the specimen is in the linear

Fig. 24. Analysis of the fracture process of three-dimensional specimens.

15
H. Wang et al. Construction and Building Materials 417 (2024) 135291

elastic stage, the stress has not reached the material strength, and the concrete directly affect its macroscopic mechanical response, and the
specimen has not broken, at which point the dissipated fracture energy is tensile strength of mortar and ITZ directly affects the change in peak
0; (b) The specimen is in the stage of crack initiation and stable stress. Moreover, the sensitivity of ITZ’s tensile strength to concrete peak
expansion, with stress exceeding the material strength, and dissipation stress is greater than that of mortar’s tensile strength, and the impact of
energy rapidly increasing while displacement slightly increasing; (c) fracture energy on concrete peak stress is limited, mainly affecting the
After the main crack penetrates, the curve shape tends to be horizontal, path of concrete cracks.
resulting in apparent cracks. The dissipated energy slowly increases (5) While paying attention to the strengthening effect of fibers on the
while the displacement rapidly increases. Adding polypropylene fibers tensile strength and fracture energy of mortar, improving the weakening
increases the dissipative fracture energy of PFRC, and the reinforcement effect of fibers on the tensile strength and fracture energy of ITZ is the
pattern is similar to the ultimate tensile strength. The dissipative frac­ key to improving the macroscopic mechanical properties of PFRC.
ture energy of concrete is not positively correlated with the fiber volume These results provide a reference for the cohesion model to simulate
ratio. When the fiber volume ratio is 0.6, the fracture dissipation energy the cracking analysis of fiber-reinforced concrete with many fibers and
is slightly higher than that of ordinary concrete. The results in Section different shapes in concrete that are not conducive to modeling.
2.4 indicate that the fracture energy of mortar is positively correlated
with the volume ratio of polypropylene fibers, while the fracture energy CRediT authorship contribution statement
of ITZ is negatively correlated with the volume ratio of polypropylene
fibers. Combining the 3D mesoscopic dissipative fracture energy curve, He Xiongjun: Writing – review & editing, Supervision, Project
the fiber volume fraction 0.4% is the turning point of dissipative fracture administration, Funding acquisition, Conceptualization. Wang Huayi:
energy. Therefore, while paying attention to the strengthening effect of Writing – original draft, Methodology, Investigation, Formal analysis,
fibers on the fracture energy of mortar, improving the weakening impact Data curation. Wei Bingyan: Data curation. Zhou Ming: Validation,
of fibers on ITZ fracture energy is the key to improving the performance Supervision, Funding acquisition. Wu Weiwei: Writing – original draft.
of PFRC. He Jia: Writing – review & editing, Supervision. Zhou Guantao: Visu­
The trend of the crack area changing with strain is consistent with alization, Supervision.
the dissipated fracture energy. As shown in Fig. 24 (b), with the increase
of fiber volume fraction, the crack area of concrete gradually increases
under uniaxial tension, resulting in a multi-crack failure mode of con­ Declaration of Competing Interest
crete. Zhou et al. [76] studied polypropylene-reinforced concrete beams
and also found that polypropylene increases the area of concrete The authors declare that they have no known competing financial
cracking, resulting in multi-crack cracking characteristics of PFRC. interests or personal relationships that could have appeared to influence
the work reported in this paper.
5. Conclusions
Data Availability
This study aims to establish an effective and efficient cracking
calculation model applicable to polypropylene fiber-reinforced concrete Data will be made available on request.
(PFRC) and to reveal the reinforcement mechanism of PFRC from a
microscopic perspective. From a microscale perspective, PFRC was Acknowledgments
viewed as a heterogeneous material consisting of aggregates, mortar
(with polypropylene fibers embedded), and the interfacial transition The authors may wish to express their sincere appreciation for the
zone (ITZ). A total of 24 mortar specimens, 24 Hybrid specimens, and 12 financial support provided by the National Natural Science Foundation
PFRC specimens were prepared with varying fiber volume fractions. of China (No. 51178361), the Science and Technology Project of the
Experimental tests were conducted to assess the influence of poly­ Department of Transportation of Hubei Province (No. 2022-11-2-8), and
propylene fibers on the mechanical properties of mortar and the mortar- the Fundamental Research Funds for the Central Universities (No. 2023-
aggregate transition zone (ITZ). A finite element model for the cracking vb-009) for this work.
of microfiber-reinforced concrete was established based on micro-scale
experiment results and CZM. The conclusion is as follows: References
(1) Adding an appropriate volume ratio of polypropylene fibers can
improve the mechanical properties of mortar, increase flexural and [1] A. Noufid, S. Belattar, Improvement of the mechanical properties of concretes and
tensile strength by 9.6–15.0% and 9.5–21.9%, and increase fracture mortars by addition of polypropylene, J. Eng. Technol. (2017).
[2] Y. Zhang, S. Zhang, M. Deng, Four-point bending tests of ECC: Mechanical response
energy by a maximum of 88.7%. and toughness evaluation, Case Stud. Constr. Mater. 17 (2022) e1573, https://doi.
(2) The hydrophobicity of PP fibers causes bubbles to adhere to the org/10.1016/j.cscm.2022.e01573.
surface of the fibers, limiting bubble elimination and increasing the local [3] S. Naderi, M. Zhang, 3D meso-scale modelling of tensile and compressive fracture
behaviour of steel fibre reinforced concrete, Compos. Struct. 291 (2022) 115690,
porosity of ITZ, thereby reducing the interfacial bonding strength be­ https://doi.org/10.1016/j.compstruct.2022.115690.
tween mortar and aggregate and weakening the mechanical properties [4] M. Zhou, X. He, H. Wang, C. Wu, J. He, Mesoscale modeling of polypropylene fiber
of ITZ. The larger the volume ratio of PP fibers, the more significant the reinforced concrete under split tension using discrete element method, Constr.
Build. Mater. 404 (2023) 133274, https://doi.org/10.1016/j.
weakening effect. The maximum flexural strength, tensile strength, and conbuildmat.2023.133274.
fracture energy after weakening are only 74.3%, 79.1%, and 37.6% of [5] P. Smarzewski, Influence of basalt-polypropylene fibres on fracture properties of
those without doped fibers. high performance concrete, Compos. Struct. 209 (2019) 23–33, https://doi.org/
10.1016/j.compstruct.2018.10.070.
(3) Although 3D and 2D models have similarities in simulating crack [6] Z. Pan, C. Wu, J. Liu, W. Wang, J. Liu, Study on mechanical properties of cost-
morphology and failure morphology, they exhibit significant differences effective polyvinyl alcohol engineered cementitious composites (PVA-ECC), Constr.
in computational mechanical responses. The stress-strain trend calcu­ Build. Mater. 78 (2015) 397–404, https://doi.org/10.1016/j.
conbuildmat.2014.12.071.
lated by the 3D model is consistent with the measured values, with a
[7] H.Z.H. Zhang, L.W.L. Wang, K.Z.K. Zheng, B.J.B.T. Jibrin, P.T.P.G. Totakhil,
stress peak-to-experimental ratio of 0.997. The 2D model differs Research on compressive impact dynamic behavior and constitutive model of
significantly from the measured values in stress-strain trends and stress polypropylene fiber reinforced concrete, Constr. Build. Mater. (2018) 584–595,
peaks, with a stress peak to experimental ratio of 0.706. The simulation https://doi.org/10.1016/j.conbuildmat.2018.07.164.
[8] W. Yang, Z. Tang, W. Wu, K. Zhang, J. Yuan, H. Li, Z. Feng, Effect of different fibers
effect of the 3D model is closer to reality. on impermeability of steam cured recycled concrete, Constr. Build. Mater. 328
(4) Whether it is a 3D or a 2D model, the cohesion parameters of (2022) 127063, https://doi.org/10.1016/j.conbuildmat.2022.127063.

16
H. Wang et al. Construction and Building Materials 417 (2024) 135291

[9] A.A. Ramezanianpour, M. Esmaeili, S.A. Ghahari, M.H. Najafi, Laboratory study on transition zone of concrete with mineral admixtures and polymer latex, Constr.
the effect of polypropylene fiber on durability, and physical and mechanical Build. Mater. 168 (2018) 946–957, https://doi.org/10.1016/j.
characteristic of concrete for application in sleepers, Constr. Build. Mater. 44 conbuildmat.2018.02.205.
(2013) 411–418, https://doi.org/10.1016/j.conbuildmat.2013.02.076. [34] C. Li, J. Guo, Cracking simulation of asphalt concrete beam specimen using
[10] C. Wu, X. He, X. Zhao, L. He, Y. Song, X. Zhang, Effect of fiber content on cohesive zone model, Constr. Build. Mater. 214 (2019) 49–60, https://doi.org/
mechanical properties and microstructural characteristics of alkali resistant glass 10.1016/j.conbuildmat.2019.04.122.
fiber reinforced concrete, Adv. Mater. Sci. Eng. 2022 (2022) 1–19, https://doi.org/ [35] H. Ji, X. Yang, Z. Luo, Experimental study of fracture properties of interfacial
10.1155/2022/1531570. transition zone by three-point bending tests on granite-mortar hybrid beams,
[11] D. Wang, Y. Ju, H. Shen, L. Xu, Mechanical properties of high performance Constr. Build. Mater. 263 (2020) 120605, https://doi.org/10.1016/j.
concrete reinforced with basalt fiber and polypropylene fiber, Constr. Build. Mater. conbuildmat.2020.120605.
197 (2019) 464–473, https://doi.org/10.1016/j.conbuildmat.2018.11.181. [36] X. Wang, S. Dong, A. Ashour, W. Zhang, B. Han, Effect and mechanisms of
[12] M.K. Yew, H.B. Mahmud, B.C. Ang, M.C. Yew, Influence of different types of nanomaterials on interface between aggregates and cement mortars, Constr. Build.
polypropylene fibre on the mechanical properties of high-strength oil palm shell Mater. 240 (2020) 117942, https://doi.org/10.1016/j.conbuildmat.2019.117942.
lightweight concrete, Constr. Build. Mater. 90 (2015) 36–43, https://doi.org/ [37] E. Schlangen, J.G.M. Van Mier, Micromechanical analysis of fracture of concrete,
10.1016/j.conbuildmat.2015.04.024. Int. J. Damage Mech. (NO.4) (1992) 435–454.
[13] L. Chen, X. Zhang, G. Liu, Analysis of dynamic mechanical properties of sprayed [38] Z.T.A.Y. Lu, Mesoscale modelling of concrete for static and dynamic response
fiber-reinforced concrete based on the energy conversion principle, Constr. Build. analysis Part 1: model development and implementation, Struct. Eng. Mech. (No.2)
Mater. 254 (2020) 119167, https://doi.org/10.1016/j.conbuildmat.2020.119167. (2011) 197–213.
[14] H. Hosseinzadeh, A. Masoud Salehi, M. Mehraein, G. Asadollahfardi, The effects of [39] J. Huang, Q. Peng, X. Hu, Y. Du, A combined-alpha-shape-implicit-surface
steel, polypropylene, and high-performance macro polypropylene fibers on approach to generate 3D random concrete mesostructures via digital image
mechanical properties and durability of high-strength concrete, Constr. Build. processing, spectral representation, and point cloud, Constr. Build. Mater. 143
Mater. 386 (2023) 131589, https://doi.org/10.1016/j.conbuildmat.2023.131589. (2017) 330–365, https://doi.org/10.1016/j.conbuildmat.2017.03.104.
[15] P. Ye, Z. Chen, W. Su, Mechanical properties of fully recycled coarse aggregate [40] H. Chen, B. Xu, Y.L. Mo, T. Zhou, Behavior of meso-scale heterogeneous concrete
concrete with polypropylene fiber, Case Stud. Constr. Mater. 17 (2022) e1352, under uniaxial tensile and compressive loadings, Constr. Build. Mater. 178 (2018)
https://doi.org/10.1016/j.cscm.2022.e01352. 418–431, https://doi.org/10.1016/j.conbuildmat.2018.05.052.
[16] Z. Yuan, Y. Jia, Mechanical properties and microstructure of glass fiber and [41] S. Naderi, M. Zhang, Meso-scale modelling of static and dynamic tensile fracture of
polypropylene fiber reinforced concrete: An experimental study, Constr. Build. concrete accounting for real-shape aggregates, Cem. Concr. Compos. 116 (2021)
Mater. 266 (2021) 121048, https://doi.org/10.1016/j.conbuildmat.2020.121048. 103889, https://doi.org/10.1016/j.cemconcomp.2020.103889.
[17] X. Zhou, Y. Zeng, P. Chen, Z. Jiao, W. Zheng, Mechanical properties of basalt and [42] O. Yılmaz, J. Molinari, A mesoscale fracture model for concrete, Cem. Concr. Res.
polypropylene fibre-reinforced alkali-activated slag concrete, Constr. Build. Mater. 97 (2017) 84–94, https://doi.org/10.1016/j.cemconres.2017.03.014.
269 (2021) 121284, https://doi.org/10.1016/j.conbuildmat.2020.121284. [43] W. Trawiński, J. Tejchman, J. Bobiński, A three-dimensional meso-scale modelling
[18] F. Xu, S. Wang, T. Li, B. Liu, B. Li, Y. Zhou, Mechanical properties and pore of concrete fracture, based on cohesive elements and X-ray μCT images, Eng. Fract.
structure of recycled aggregate concrete made with iron ore tailings and Mech. 189 (2018) 27–50, https://doi.org/10.1016/j.engfracmech.2017.10.003.
polypropylene fibers, J. Build. Eng. 33 (2021) 101572, https://doi.org/10.1016/j. [44] X.X.T.E. Xiong, Q. Xiao, Meso-Scale Simulation of Concrete Based on Fracture and
jobe.2020.101572. Interaction Behavior, Appl. Sci. (No.15) (2019) 2986, https://doi.org/10.3390/
[19] H. Guo, J. Tao, Y. Chen, D. Li, B. Jia, Y. Zhai, Effect of steel and polypropylene app9152986.
fibers on the quasi-static and dynamic splitting tensile properties of high-strength [45] W. Ren, Z. Yang, R. Sharma, C. Zhang, P.J. Withers, Two-dimensional X-ray CT
concrete, Constr. Build. Mater. 224 (2019) 504–514, https://doi.org/10.1016/j. image based meso-scale fracture modelling of concrete, Eng. Fract. Mech. 133
conbuildmat.2019.07.096. (2015) 24–39, https://doi.org/10.1016/j.engfracmech.2014.10.016.
[20] S. Fallah, M. Nematzadeh, Mechanical properties and durability of high-strength [46] X. Wang, M. Zhang, A.P. Jivkov, Computational technology for analysis of 3D
concrete containing macro-polymeric and polypropylene fibers with nano-silica meso-structure effects on damage and failure of concrete, Int. J. Solids Struct. 80
and silica fume, Constr. Build. Mater. 132 (2017) 170–187, https://doi.org/ (2016) 310–333, https://doi.org/10.1016/j.ijsolstr.2015.11.018.
10.1016/j.conbuildmat.2016.11.100. [47] JTG E42–2005, Test Methods of Aggregate for Highoay Engineering, China
[21] V. Afroughsabet, T. Ozbakkaloglu, Mechanical and durability properties of high- Communications Press, 2005.
strength concrete containing steel and polypropylene fibers, Constr. Build. Mater. [48] JTG/T 3650-2020, Technical Specifications for Construction of Highway Bridges
94 (2015) 73–82, https://doi.org/10.1016/j.conbuildmat.2015.06.051. and Culverts, Chinese Standard Press, 2020.
[22] A. Hosan, F.U.A. Shaikh, P. Sarker, F. Aslani, Nano- and micro-scale [49] GB/T 14337-2022, Man-made Fibre-Test Method for Tensile Properties of Staple
characterisation of interfacial transition zone (ITZ) of high volume slag and slag-fly Fibre, Chinese Standard Press, 2022.
ash blended concretes containing nano SiO2 and nano CaCO3, Constr. Build. [50] D. Altalabani, D.K.H. Bzeni, S. Linsel, Mechanical properties and load deflection
Mater. 269 (2021) 121311, https://doi.org/10.1016/j.conbuildmat.2020.121311. relationship of polypropylene fiber reinforced self-compacting lightweight
[23] Y. Shi, X. Lv, S. Zhou, Z. Liu, M. Yang, C. Liu, C. Lu, Mechanical properties, concrete, Constr. Build. Mater. 252 (2020) 119084, https://doi.org/10.1016/j.
durability, and ITZ characteristics of full-grade dam concrete prepared by conbuildmat.2020.119084.
aggregates with surface rust stains, Constr. Build. Mater. 305 (2021) 124798, [51] D. Zheng, W. Song, J. Fu, G. Xue, J. Li, S. Cao, Research on mechanical
https://doi.org/10.1016/j.conbuildmat.2021.124798. characteristics, fractal dimension and internal structure of fiber reinforced concrete
[24] K. Wu, H. Han, H. Li, B. Dong, T. Liu, G. De Schutter, Experimental study on under uniaxial compression, Constr. Build. Mater. 258 (2020) 120351, https://doi.
concurrent factors influencing the ITZ effect on mass transport in concrete, Cem. org/10.1016/j.conbuildmat.2020.120351.
Concr. Compos. 123 (2021) 104215, https://doi.org/10.1016/j. [52] N. Ranjbar, M. Mehrali, M. Mehrali, U.J. Alengaram, M.Z. Jumaat, Graphene
cemconcomp.2021.104215. nanoplatelet-fly ash based geopolymer composites, Cem. Concr. Res. 76 (2015)
[25] R. Zhou, Z. Song, Y. Lu, 3D mesoscale finite element modelling of concrete, 222–231, https://doi.org/10.1016/j.cemconres.2015.06.003.
Comput. Struct. 192 (2017) 96–113, https://doi.org/10.1016/j. [53] H. Mazaheripour, S. Ghanbarpour, S.H. Mirmoradi, I. Hosseinpour, The effect of
compstruc.2017.07.009. polypropylene fibers on the properties of fresh and hardened lightweight self-
[26] P. Wriggers, S.O. Moftah, Mesoscale models for concrete: Homogenisation and compacting concrete, Constr. Build. Mater. 25 (1) (2011) 351–358, https://doi.
damage behaviour, Finite Elem. Anal. Des. 42 (7) (2006) 623–636, https://doi. org/10.1016/j.conbuildmat.2010.06.018.
org/10.1016/j.finel.2005.11.008. [54] GB/T 50081-2019, Standard for Test Methods of Concrete Physical and Mechanical
[27] S. Naderi, M. Zhang, A novel framework for modelling the 3D mesostructure of Properties, Chinese Standard Press, 2019.
steel fibre reinforced concrete, Comput. Struct. 234 (2020) 106251, https://doi. [55] GB/T 50082-2009, Standard for Test Methods of Long-term Performance and
org/10.1016/j.compstruc.2020.106251. Durability of Ordinary Concrete, Chinese Standard Press, 2009.
[28] J. Wang, X. Li, A.P. Jivkov, Q.M. Li, D.L. Engelberg, Interfacial transition zones in [56] K. Hannawi, H. Bian, W. Prince-Agbodjan, B. Raghavan, Effect of different types of
concrete meso-scale models – Balancing physical realism and computational fibers on the microstructure and the mechanical behavior of ultra-high
efficiency, Constr. Build. Mater. 293 (2021) 123332, https://doi.org/10.1016/j. performance fiber-reinforced concretes, Compos. Part B: Eng. 86 (2016) 214–220,
conbuildmat.2021.123332. https://doi.org/10.1016/j.compositesb.2015.09.059.
[29] Y. Zhang, Q. Chen, Z. Wang, J. Zhang, Z. Wang, Z. Li, 3D mesoscale fracture [57] Q. Yu, H. Liu, T. Yang, H. Liu, 3D numerical study on fracture process of concrete
analysis of concrete under complex loading, Eng. Fract. Mech. 220 (2019) 106646, with different ITZ properties using X-ray computerized tomography, Int. J. Solids
https://doi.org/10.1016/j.engfracmech.2019.106646. Struct. 147 (2018) 204–222, https://doi.org/10.1016/j.ijsolstr.2018.05.026.
[30] L. Jin, J. Li, W. Yu, X. Du, Mesoscopic simulations on the strength and size effect of [58] T. Liu, S. Qin, D. Zou, W. Song, J. Teng, Mesoscopic modeling method of concrete
concrete under biaxial loading, Eng. Fract. Mech. 253 (2021) 107870, https://doi. based on statistical analysis of CT images, Constr. Build. Mater. 192 (2018)
org/10.1016/j.engfracmech.2021.107870. 429–441, https://doi.org/10.1016/j.conbuildmat.2018.10.136.
[31] M. Maleki, I. Rasoolan, A. Khajehdezfuly, A.P. Jivkov, On the effect of ITZ [59] L. Shuguang, L. Qingbin, Method of meshing ITZ structure in 3D meso-level finite
thickness in meso-scale models of concrete, Constr. Build. Mater. 258 (2020) element analysis for concrete, Finite Elem. Anal. Des. 93 (2015) 96–106, https://
119639, https://doi.org/10.1016/j.conbuildmat.2020.119639. doi.org/10.1016/j.finel.2014.09.006.
[32] H. Qiu, Z. Zhu, M. Wang, F. Wang, C. Luo, D. Wan, Study of the failure properties [60] H. Xie, J. Feng, Implementation of numerical mesostructure concrete material
and tensile strength of rock-mortar interface transition zone using bi-material models: a dot matrix method, Materials 12 (23) (2019) 3835, https://doi.org/
Brazilian discs, Constr. Build. Mater. 236 (2020) 117551, https://doi.org/ 10.3390/ma12233835.
10.1016/j.conbuildmat.2019.117551. [61] C. Du, L. Sun, S. Jiang, Z. Ying, Numerical simulation of aggregate shapes of three-
[33] A. Qudoos, Atta-ur-Rehman, H.G. Kim, J.S. Ryou, Influence of the surface dimensional concrete and its applications, J. Aerosp. Eng. 26 (3) (2013) 515–527,
roughness of crushed natural aggregates on the microhardness of the interfacial https://doi.org/10.1061/(ASCE)AS.1943-5525.0000181.

17
H. Wang et al. Construction and Building Materials 417 (2024) 135291

[62] G. Zhou, Z. Xu, 3D mesoscale investigation on the compressive fracture of concrete [69] M.K.M.L. Benzeggagh, Measurement of mixed-mode delamination fracture
with different aggregate shapes and interface transition zones, Constr. Build. toughness of unidirectional glass/epoxy composites with mixed-mode bending
Mater. 393 (2023) 132111, https://doi.org/10.1016/j.conbuildmat.2023.132111. apparatus, Compos. Sci. Technol. (No.4) (1996) 439–449.
[63] P.S.M. Thilakarathna, K.S.K. Baduge, P. Mendis, V. Vimonsatit, H. Lee, Mesoscale [70] S. Naderi, W. Tu, M. Zhang, Meso-scale modelling of compressive fracture in
modelling of concrete - a review of geometry generation, placing algorithms, concrete with irregularly shaped aggregates, Cem. Concr. Res. 140 (2021) 106317,
constitutive relations and applications, Eng. Fract. Mech. 231 (2020) 106974, https://doi.org/10.1016/j.cemconres.2020.106317.
https://doi.org/10.1016/j.engfracmech.2020.106974. [71] H.W. Reinhardt, S. Xu, A practical testing approach to determine mode II fracture
[64] P. Hajikarimi, T.F. Fakhari, N.F. Moghadas, J. Absi, M. Rahi, A. Khodaii, C. Petit, energy G/sub IIF/ for concrete, Int. J. Fract. (No.2) (2000) 107–125.
Mechanical behavior of polymer-modified bituminous mastics. I: experimental [72] Z.P. Bažant, P.A. Pfeiffer, Shear fracture tests of concrete, Mater. Struct. (No.2)
approach, J. Mater. Civ. Eng. 31 (1) (2019) 4018337, https://doi.org/10.1061/ (1986) 111–121.
(ASCE)MT.1943-5533.0002548. [73] C. Naga Satish Kumar, T.D. Gunneswara Rao, Punching shear resistance of concrete
[65] X. Qin, C. Gu, C. Shao, X. Fu, L. Vallejo, Y. Chen, Numerical analysis of fracturing slabs using mode-II fracture energy, Eng. Fract. Mech. 83 (2012) 75–85, https://
behavior in fully-graded concrete with oversized aggregates from mesoscopic doi.org/10.1016/j.engfracmech.2012.01.004.
perspective, Constr. Build. Mater. 253 (2020) 119184, https://doi.org/10.1016/j. [74] X.F. Wang, Z.J. Yang, J.R. Yates, A.P. Jivkov, C. Zhang, Monte Carlo simulations of
conbuildmat.2020.119184. mesoscale fracture modelling of concrete with random aggregates and pores,
[66] X. Chen, J. Yuan, Q. Dong, X. Zhao, Meso-scale cracking behavior of cement treated Constr. Build. Mater. 75 (2015) 35–45, https://doi.org/10.1016/j.
base material, Constr. Build. Mater. 239 (2020) 117823, https://doi.org/10.1016/ conbuildmat.2014.09.069.
j.conbuildmat.2019.117823. [75] S. Gao, Nonlinear finite element failure analysis of bolted steel-concrete composite
[67] G. Alfano, On the influence of the shape of the interface law on the application of frame under column-loss, J. Constr. Steel Res. 155 (2019) 62–76, https://doi.org/
cohesive-zone models, Compos. Sci. Technol. 66 (6) (2006) 723–730, https://doi. 10.1016/j.jcsr.2018.12.020.
org/10.1016/j.compscitech.2004.12.024. [76] J.Z.J. Zhou, W.S.W. Shen, S.W.S. Wang, Experimental study on torsional behavior
[68] Z.J. Yang, X.T. Su, J.F. Chen, G.H. Liu, Monte Carlo simulation of complex cohesive of FRC and ECC beams reinforced with GFRP bars, Constr. Build. Mater. (2017)
fracture in random heterogeneous quasi-brittle materials, Int. J. Solids Struct. 46 74–81, https://doi.org/10.1016/j.conbuildmat.2017.06.131.
(17) (2009) 3222–3234, https://doi.org/10.1016/j.ijsolstr.2009.04.013.

18